Research Article
A Gabor-Block-Based Kernel Discriminative Common Vector Approach Using Cosine Kernels for Human Face Recognition
Table 3
Average recognition results using FERET database.
| | | Recognition rates (%) | | | Method | No. of training samples | Average recognition rates (%) | | | 3 | 4 | |
| | GWT | 79.65 | 82.5 | 81.1 | | KDCV | 69.5 | 75 | 72.25 | | GWT-LDA | 82.76 | 85.33 | 84.1 | | GWT-KDCV (RBF) | 81 | 82.50 | 81.75 | | GWT-KDCV (Cosine) | 83 | 85 | 84.5 | | GWT-KDCV (RBF, low-energized) | 88.5 | 91.5 | 90 | | GWT-KDCV (Cosine, low-energized) | 92.5 | 95.75 | 94.1 |
|
|
